On January 28, 2020, we released our sixth webinar, Enrichment Trials. This topic is presented by Dr. Kert Viele of Berry Consultants, LLC.
Very few therapies benefit everyone. When we have uncertainty about “who benefits?”, a standard “one size fits all” clinical trial will fail. If, for example, half the population benefits and the other half does not, then any overall conclusion will miss half the population. If the trial is successful, we may mistakenly give the treatment to the poorly treated half. If the trial is unsuccessful, often this is because of the dilution of the treatment effect from the poorly treated half. Post-hoc subgroup analyses attack this problem, but are typically underpowered and prone to multiplicity issues.
Enrichment trials prospectively investigate subgroups. During the trial, we estimate the effect of our therapy in each subgroup, and adaptively drop subgroups that are performing poorly, allowing for greater sample sizes in the remaining subgroups. These greater sample sizes allow for increased power in the presence of differential subgroup effects. We will review these methodologies, discuss their benefits and risks (for example, they are typically larger than a standard trial), and provide examples from the literature.